Organization and Regulation of Pentachlorophenol-Degrading Genes in Sphingobium chlorophenolicum ATCC 39723

Abstract

The first three enzymes of the pentachlorophenol (PCP) degradation pathway in Sphingobium chlorophenolicum (formerly Sphingomonas chlorophenolica) ATCC 39723 have been characterized, and the corresponding genes, pcpA, pcpB, and pcpC, have been individually cloned and sequenced. To search for new genes involved in PCP degradation and map the physical locations of the pcp genes, a 24-kb fragment containing pcpA and pcpC was completely sequenced. A putative LysR-type transcriptional regulator gene, pcpM, and a maleylacetate reductase gene, pcpE, were identified upstream of pcpA. pcpE was found to play a role in PCP degradation. pcpB was not found on the 24-kb fragment. The four gene products PcpB, PcpC, PcpA, and PcpE were responsible for the metabolism of PCP to 3-oxoadipate in ATCC 39723, and inactivational mutation of each gene disrupted the degradation pathway. The organization of the pcp genes is unusual because the four PCP-degrading genes, pcpA, pcpB, pcpC, and pcpE, were found to be located at four discrete locations. Two hypothetical LysR-type regulator genes, pcpM and pcpR, have been identified; pcpM was not required, but pcpR was essential for the induction of pcpB, pcpA, and pcpE. The coinducers of PcpR were PCP and other polychlorinated phenols. The expression of pcpC was constitutive. Thus, the organization and regulation of the genes involved in PCP degradation to 3-oxoadipate were documented.

Pentachlorophenol (PCP) has been released into the environment as a wood preservative (8, 13). This compound is a major environmental pollutant due to its toxicity and recalcitrance, and it is regulated as one of the priority pollutants by the U.S. Environmental Protection Agency (16, 30). Microorganisms have been used to remove PCP from the environment (16, 17), and several aerobic PCP-degrading bacteria have been isolated from contaminated soils (7). Sphingobium chlorophenolicum (31) (formerly Sphingomonas chlorophenolica) strain ATCC 39723 is one of the bacteria capable of completely mineralizing PCP (24). The biochemistry of PCP degradation by ATCC 39723 has been extensively studied (Fig. 1). PCP 4-monooxygenase (PcpB) oxidizes PCP to 2,3,5,6-tetrachloro-p-hydroquinone (TeCH) (22, 35, 37, 38). TeCH reductive dehalogenase (PcpC) converts TeCH to 2,3,6-trichloro-p-hydroquinone and then to 2,6-dichloro-p-hydroquinone (DiCH) by reductive dechlorination (20, 39, 40). DiCH is subject to ring cleavage by DiCH 1,2-dioxygenase (PcpA), producing 2-chloromaleylacetate (2-CMA) (19, 33). The corresponding genes, pcpB, pcpC, and pcpA, have been individually cloned and sequenced (21, 22, 36). pcpB was found to be physically linked with two other putative pcp genes, pcpD and pcpR (21). pcpR is a hypothetical LysR-type regulator. Northern hybridization and enzymatic activity analysis suggest that PcpB and PcpA are PCP inducible in strain ATCC 39723 (22, 33), while PcpC is constitutively produced (20, 40). However, the overall organization and regulation of PCP-degrading genes have not been reported, and the metabolic steps beyond ring cleavage by PcpA have not been characterized.

FIG. 1.

PCP degradation pathway in S. chlorophenolicum strain ATCC 39723. PcpB, PCP 4-monooxygenase; PcpC, TeCH reductive dehalogenase; PcpA, 2,6-DiCH 1,2-dioxygenase; PcpE, MA reductase; CoA, coenzyme A.

We report here complete sequencing of two DNA fragments. A 24-kb DNA fragment contained pcpA and pcpC, as well as two new pcp genes, pcpE encoding a maleylacetate (MA) reductase and pcpM coding for a hypothetical LysR-type regulator. An 8-kb DNA fragment containing the pcpB-pcpD-pcpR cluster (21) was also sequenced. Sequence data revealed the discrete organization of each functional pcp gene; genetic and functional analyses of pcpE, pcpC, pcpM, and pcpR documented the PCP degradation pathway to 3-oxoadipate and the regulation of PCP-degrading genes in ATCC 39723.

MATERIALS AND METHODS

Bacterial strains, plasmids, and culture conditions.

S. chlorophenolicum ATCC 39723, its pcpA mutant (4), and its pcpB mutant (Fco315) (15) were generous gifts from Ronald L. Crawford and Cindy S. Orser. Escherichia coli DH5απ and the vector pSG76-K (23) were generous gifts from G. Pósfai. Subcloning of DNA fragments for sequencing was done in pBluescript II KS(+) (Stratagene, La Jolla, Calif.). A glutamate (0.4%) mineral salt (GMS) medium was used to grow ATCC 39723 at 30°C (24). Kanamycin was added to GMS medium at a concentration of 5 μg · ml⁻¹ when it was needed. E. coli strains were grown in Luria-Bertani medium (25). Ampicillin or kanamycin was added at a concentration of 100 or 30 μg · ml⁻¹ to Luria-Bertani medium when required.

DNA preparation and manipulation.

Plasmid DNA isolation was carried out with a Miniprep or Midiprep kit (Qiagen, Chatsworth, Calif.). DNA digestion, electrophoresis, ligation, and transformation were performed by using standard procedures (25). A Geneclean kit (Bio 101, Carlsbad, Calif.) was used for recovery of DNA fragments from agarose gels. For colony PCR, a portion of a colony was directly added to a PCR mixture as the DNA template. Long-range PCR was done with an XL PCR kit (Perkin-Elmer, Branchburg, N.J.).

TAIL-PCR.

TAIL-PCR (18) was performed to amplify unknown DNA sequences contiguous to known pcp gene sequences. Specific primer TAIL1 and an arbitrary primer were used in a first round of long-range PCR with ATCC 39723 genomic DNA as the template and an annealing temperature of 55°C. Products from the first round of PCR were used as templates for a second round of long-range PCR performed with the same arbitrary primer and another specific primer, TAIL2, about 200 bp downstream of primer TAIL1; the annealing temperature used was 62°C. Products from the first and second rounds of PCR were separated on an agarose gel for comparison. Products showing the expected decrease in size after the second round of PCR were gel purified and directly sequenced with primer TAIL2.

DNA sequencing and sequence analysis.

DNA cycle sequencing reactions with the ABI Prism Big Dye-labeled dideoxynucleotides (Applied Biosystems, Foster City, Calif.) were used for DNA sequencing. The GCG programs (Genetics Computer Group, Madison, Wis.) were used for compiling sequences and for analyses of the terminator, codon usage, translation, G+C content, and homology. DNA and deduced protein sequences were searched for similarities to database entries by using BLAST programs available through the National Center for Biotechnology Information website (http://www.ncbi.nlm.nih.gov/BLAST/).

Inactivation of pcpE, pcpC, pcpR, and pcpM.

An internal fragment of the target gene was PCR amplified with specific primers for each gene (Table 1). KpnI and PstI sites were introduced into the forward and reverse primers, respectively. The resultant pcpE_inter, pcpC_inter, pcpR_inter, or pcpM_inter PCR product was cloned into the suicide vector pSG76-K at KpnI/PstI sites (23). The recombinant plasmid DNA was electroporated into E. coli DH5απ for maintenance and recovery. The identity of the recovered plasmid was confirmed by sequencing, and then the plasmid was electroporated into ATCC 39723 by using a modified method of Lange et al. (15). Fifty microliters of ATCC 39723 competent cells was mixed with ca. 100 ng of the plasmid, and the mixture was pulsed three times at 2.5 kV by using a Gene Pulser (Bio-Rad, Hercules, Calif.). One milliliter of GMS medium containing 1 mM MgCl₂ and 0.3 mM CaCl₂ was immediately added to the cells, and the cells were incubated with shaking for 6 h at 30°C before they were plated on GMS agar containing kanamycin. Mutants were identified by their kanamycin resistance and were confirmed by colony PCR by using a pair of primers flanking the internal sequence (Table 1).

TABLE 1.

Oligonucleotide primers used for PCR in this study

Primer	Sequencea	Target gene	Nucleotide positionb
Primers for amplification of internal     fragments of target genes
EF-1	5′-AAGGCGAAGGTACCGGACGTCGAC-3′	pcpE	11331-11308
ER-1	5′-TTCATAGAGCTGCAGCGCCGGATC-3′	pcpE	10687-10710
RF-1	5′-CATTTCGACGGTACCCTGTTCGTG-3′	pcpR	3429-3452
RR-1	5′-CACCTGCTCCTGCAGATGATTGAA-3′	pcpR	3923-3900
MF-1	5′-ACGCTCAGCCGGTACCTGAAGCAG-3′	pcpM	14689-14712
MR-1	5′-ATATGGGCGCTCTGCAGGAAGCTG-3′	pcpM	14111-14134
CF-1	5′-CTGTTCGATGGGTACCCGTCTGGC-3′	pcpC	21847-21824
CR-1	5′-CCAGATGCACCTGCAGTTCGTCGA-3′	pcpC	21339-21362
Primers for confirmation of gene disruption
EF-2	5′-GAGAGGAATCATATGCAGTTCGTT-3′	pcpE	11574-11551
ER-2	5′-GGATACAAGCTTCTATATGGTGAC-3′	pcpE	10492-10515
RF-2	5′-GGAACCGCATATGAATGATTCCGT-3′	pcpR	3281-3304
RR-2	5′-GATGCCGCGCAAGCTTCTCCGCCGGCGGAT-3′	pcpR	4227-4198
MF-2	5′-AGGAAATGACCATATGATATCCTT-3′	pcpM	14818-14795
MR-2	5′-CGCTCAATAAGCTTGTCTGGCGCG-3′	pcpM	13871-13894
CF-2	5′-CGCTGAGGATATGAAGATGCCTGA-3′	pcpC	21901-21878
CR-2	5′-CATCCGCTGATAATAAGCGAGCAG-3′	pcpC	21199-21222
Primers for RT-PCR analysis
BF-1	5′-CGGGTTCACGTTCAACTTCGAGAA-3′	pcpB	579-602
BR-1	5′-GATCGTCGAAGGAACTGAGATAGC-3′	pcpB	1126-1103
CF-3	5′-CTATGACGACAAGCAGGTGGACAT-3′	pcpC	21802-21779
CR-3	5′-CATCCGCTGATAATAAGCGAGCAG-3′	pcpC	21199-21222
AF-1	5′-CGAACCATATCACCAGTCTGCATC-3′	pcpA	15028-15051
AR-1	5′-CATGAAGAAGTCCATGTCCTCCAG-3′	pcpA	15545-15522
EF-3	5′-TCCATATCGGGTTATCTTCGGTCC-3′	pcpE	11539-11516
ER-3	5′-ATCGGGATCGTAGACCACGATCTT-3′	pcpE	11107-11130
RF-3	5′-ACGTCGAACCGGAATTTCCGCATA-3′	pcpR	3570-3593
RR-3	5′-ACTCGATCGGAACCGGAAATATGG-3′	pcpR	4098-4075
MF-3	5′-CCAGACGCTGGAACTTTATCACAG-3′	pcpM	14596-14573
MR-3	5′-TGGTCAGGATGAGTTCCGACTGTT-3′	pcpM	14082-14105

^aBoldface type indicates changes made for the introduction of restriction sites (underlined) in primers.

^bThe accession numbers for pcpEMAC and pcpBR are AF512952 and U12290, respectively.

Whole-cell PCP degradation.

PCP degradation by whole cells was monitored spectrophotometrically (34) or by a previously described high-performance liquid chromatography (HPLC) method (40). ATCC 39723 and the mutants of this strain were grown in 200 ml of GMS medium at 30°C with shaking. When the optical density at 600 nm of a cell culture reached between 0.3 and 0.4, PCP from a stock solution (37.6 mM PCP in 2.5% NaOH) was added to a final concentration of 150 μM. Incubation was continued with shaking, and samples (1 ml) were taken periodically to measure the cell density and PCP concentration. Supernatants of the PCP-degrading cultures were also analyzed for the accumulation of TeCH, DiCH, and 2-CMA by using previously described HPLC methods (29, 40). 2-CMA was prepared by using PcpA to convert DiCH to 2-CMA (33).

RT-PCR analysis.

Total RNA was prepared as previously described (36). All RNA samples were free of DNA, as confirmed by using 200 ng of isolated RNA in a 100-μl PCR mixture. A reverse transcription (RT)-PCR was carried out by using a OneStep RT-PCR kit (Gibco BRL, Gaithersburg, Md.) and a 100-μl reaction mixture with 2 ng of RNA. The primers used for RT-PCR were sequence specific for each pcp gene (Table 1). All samples were amplified by using 30 PCR cycles.

Enzyme assays with cell extracts of S. chlorophenolicum strains.

Cells that were grown in GMS medium and either induced by PCP (75 μM) or not induced were harvested by centrifugation. Cell extracts were obtained by using a French pressure cell as previously described (33) and were stored at −80°C. Protein concentrations were determined with a dye reagent (Bio-Rad) by using bovine serum albumin as a standard. The activities of PcpB, PcpC, and PcpA were determined by previously described methods (33, 35, 40). PcpE (2-CMA reductase) activity was determined spectrophotometrically by monitoring NADH oxidation at 340 nm in the presence of 2-CMA in 20 mM potassium phosphate buffer (pH 7.0) at 24°C (27). 2-CMA reductase activities were also monitored by using HPLC to monitor the consumption of 2-CMA and MA (29).

Nucleotide sequence accession numbers.

The nucleotide sequences of pcpEMAC and pcpBDR fragments have been deposited in the GenBank database under accession no. AF512952 and U12290, respectively.

RESULTS

Identification of pcpM and pcpE.

pcpA is located on pLX 001, a cosmid clone of a BamHI fragment from ATCC 39723 (36). After the BamHI fragment was subcloned into pBluescript II KS(+) as EcoRI or EcoRI/BamHI fragments, the inserts were sequenced. Outwardly oriented primers were then designed adjacent to EcoRI sites of each insert and used in PCR performed with the genomic DNA as the template in order to confirm the physical linkage of two inserts at an EcoRI site. The sequence of linked DNA fragments was assembled into a 12-kb BamHI fragment. Both pcpA and pcpC were on the BamHI fragment (Fig. 2A). A hypothetical LysR-type transcriptional regulatory gene, designated pcpM, was identified directly upstream of pcpA; the deduced amino acid sequence was most similar to the sequence of PcpR (GenBank accession no. U12290), with 43% identity. In addition, a partial open reading frame (ORF) possibly coding for an MA reductase was identified at the 5′ end of the 12-kb BamHI fragment. An approximately 10-kb DNA sequence containing the missing part of the ORF was amplified by TAIL-PCR and sequenced. The ORF was designated pcpE (Fig. 2A). PcpE was most similar to MA reductase (TftE) of Burkholderia cepacia AC1100 (9), with 50% identity in the amino acid sequence.

FIG. 2.

Gene organization of the pcpEMAC fragment and the pcpBDR fragment in ATCC 39723. Only EcoRI (E) and BamHI (B) restriction sites are indicated. The sequences of pcpA (GenBank accession no. M55159), pcpC (M98559), pcpB (M98557), and pcpDR (U12290) have been determined separately and previously deposited in the GenBank database. All other sequences were determined in this study. The solid arrows indicate the pcp genes. The potential functions of other genes are summarized in Table 2.

Organization of the pcp genes and adjacent genes.

A continuous 24-kb DNA fragment containing pcpA, pcpC, pcpM, and pcpE was assembled; this fragment included the 12-kb BamHI fragment, as well as a 10-kb product at the 5′ end and a 2-kb PCR product at the 3′ end. Sequence analysis revealed 15 additional ORFs and one partial ORF on this 24-kb fragment (Fig. 2A). pcpC, pcpA, pcpM, and pcpE were separated from each other on the fragment and were apparently transcribed from different promoters. There was a 2,323-bp region, containing two ORFs (orf11 and orf12), between pcpE and pcpM. The predicted products of orf11 and orf12 are homologous to enzymes involved in one-carbon metabolism (Table 2). pcpM and pcpA were divergently oriented with a 215-bp spacer. pcpA and pcpC were separated by 5,153 bp of DNA that included three ORFs (orf15, orf16, and orf17). These three putative genes do not have any apparent functions in PCP degradation (Table 2). pcpB was not on the 24-kb fragment, and it is part of the pcpBDR cluster (21). About 4 kb of DNA sequence 3′ to the pcpBDR cluster was also amplified by TAIL-PCR and sequenced. It contained three additional ORFs (Fig. 2B) that are not involved in biodegradation. Overlapping sequences between pcpEMAC and pcpBDR were not found.

TABLE 2.

Putative ORFs flanking the pcp genes and their predicted products

ORF	Position (nt)a	Probable function or product	Predicted mol wt	Homologous proteinb	% Identityc	Reference accession no.d
ORF1	1876-242	2,4-Dihydroxybenzoate monooxygenase	58,648	DxnD	34	CAA51370
ORF2	2599-3162	p-Nitrobenzoate reductase	20,666	PnbA	28	AAF01445
ORF3	3880-3494	Unknown	14,634	Mlr2600	57	BAB49690
ORF4	3969-4586	Acyl carrier protein phosphodiesterase	21,457	AcpD	53	NP_385442
ORF5	5889-4624	p-Aminobenzoate synthase	46,783	BMEH0013	43	AAL53254
ORF6	6196-7236	Unknown	37,830	None
ORF7	8048-7371	Glutathione S-transferase	24,476	Gst	40	CAC91172
ORF8	9530-8109	TonB-dependent receptor	50,202	Orf1217	37	AAD04012
ORF9	9824-10441	Unknown	21,747	M113682	28	BAB50526
ORF11	11902-12759	Formyltetrahydrofolate deformylase	31,768	CC3630	59	AAK25592
ORF12	12763-13713	Methylenetetrahydrofolate dehydrogenase	33,363	CC1217	51	AAK23199
ORF15	17346-16216	Na+/H+ antiporter	39,152	YjbQ	31	A69845
ORF16	18513-17536	Sulfohydrolase precursor	34,377	RSc2495	39	CAD16202
ORF17	20904-18529	TonB-dependent receptor	85,199	Orf1217	42	AAD04012
ORF19	23030-22089	Glutathione S-transferase	35,382	BME10455	55	AAL51636
ORF20	23933-23014	LysR-type regulator	Partial	LinR	33	BAA36280
ORF21	4259-5083	Succinoglycan biosynthesis protein	30,284	Smc01637	29	CAC46828
ORF22	7441-5144	Outer membrane salicin receptor	83,033	SalC	36	AAF66621
ORF23	8281-7736	TonB-dependent receptor	Partial	CC1809	28	AAK23785

^aThe accession number for ORF1 to ORF20 is AF512952, and the accession number for ORF21 to ORF23 is U12290. nt, nucleotides.

^bThe most homologous protein identified by BLASTP.

^cThe percentage of amino acid identity between two proteins was determined by using the GCG GAP program.

^dAccession number of homologous protein in EMBL/GenBanK/NCBI-RefSeq/formats.

Inactivation of pcpE, pcpC, pcpM, and pcpR in ATCC 39723.

In order to demonstrate the function of pcpE, pcpM, and pcpR and to confirm the role of pcpC in PCP degradation, the four genes were disrupted in ATCC 39723. To inactivate a target gene, an internal fragment of the gene was cloned in the suicide plasmid pSG76-K (23). Integration of the plasmid via homologous recombination between the plasmid and the genome disrupted the target gene and resulted in the formation of two truncated genes, one without the 3′ coding region and the other without the 5′ coding region. Kanamycin-resistant transformants appeared on agar plates after 7 days of incubation at 30°C. The integration was confirmed by long-range PCR performed with primers flanking the internal fragment. For example, an expected 1,083-bp product was amplified with primers EF-2 and ER-2 (Table 1) from the wild type, while a 3,492-bp product was amplified with the same primers from the mutant due to the insertion of pSG76-K::pcpE_inter inside pcpE (Fig. 3). The PCR product from the mutant was sequenced to confirm the insertion. By using this approach, pcpC, pcpM, and pcpR were also disrupted, and the mutations were confirmed.

FIG. 3.

Confirmation of pcpE disruption. The integration of an internal pcpE fragment, cloned in pSG76-K, resulted in two truncated copies of pcpE, one without the N-terminal coding region and the other without the C-terminal coding region. A pair of primers, EF-2 and ER-2 (Table 1), was designed from the missing N-terminal and C-terminal regions. The primers were used in long-range colony PCR to give a 1,083-bp predicted product with the wild-type colonies and a 3,492-bp predicted product with the pcpE mutant colonies. Lanes 1 and 4, 1-kb DNA molecular markers (Gibco); lane 2, wild type; lane 3, pcpE mutant.

PCP degradation by whole cells.

To evaluate the role of the pcp genes in PCP metabolism, ATCC 39723 and its mutants were tested for PCP degradation and the accumulation of metabolic intermediates. The wild type, growing on glutamate, took about 30 min to start degrading PCP, and 150 μM PCP was rapidly degraded (Fig. 4A). There was no apparent difference between the pcpM mutant and the wild type in terms of PCP degradation (Fig. 4A). In contrast, the pcpR mutant was unable to metabolize PCP (Fig. 4A). The pcpE mutant degraded PCP at a slightly lower rate than the wild type; however, the pcpE mutant accumulated 2-CMA but the wild type did not (Fig. 4A). Because the pcpC mutant consumed 150 μM PCP very slowly, 75 μM PCP was used to demonstrate complete consumption of PCP (Fig. 4B). During PCP degradation, TeCH, which is colorless, transitorily accumulated in the culture medium. The culture medium transitorily turned bright yellow, brown, and finally purple-pink. Gradual accumulation of a product that had an absorption peak at 347 nm was observed only in the pcpC mutant culture (Fig. 4C). When the pcpC mutant was used to degrade 2,4,6-trichlorophenol (TriCP) and 2,6-dichlorophenol (DiCP), it consumed these two chlorinated phenols without accumulation of any colored substances.

FIG. 4.

PCP degradation and intermediate accumulation by cultures of ATCC 39723 and its mutants. Cells were grown in GMS medium to an optical density at 600 nm of 0.3 to 0.4, and PCP was added to initiate degradation of PCP. It took about 30 min for the wild-type cells to begin PCP degradation. (A) PCP consumption by the wild type (▵), the pcpE mutant (♦), the pcpM mutant (×), and the pcpR mutant (○) and 2-CMA accumulation by the pcpE mutant (▪). (B) PCP consumption by the wild type (▵) and the pcpC mutant (♦) and transitory TeCH accumulation by the pcpC mutant (▪). (C) Spectral changes associated with the metabolism of PCP by the pcpC mutant. The absorption spectra of the culture supernatants (the same supernatants that were used for the experiments whose results are shown in panel B) were recorded after PCP addition at 0, 1, 2, 3, and 4 h. A product with an absorption maximum at 347 nm accumulated in the pcpC mutant culture. The wild type did not accumulate 2-CMA, TeCH, or the product with an absorption maximum at 347 nm. The whole-cell assays were repeated several times, and the trends shown were observed in each case.

RT-PCR analysis of transcription of the pcp genes in ATCC 39723 and its mutants.

In ATCC 39723 transcription of pcpB, pcpA, and pcpE was inducible by PCP, while pcpC was constitutively expressed (Fig. 5A). Expression of pcpB and pcpE was not detectable with uninduced cells, while pcpA expression increased from the basal level after PCP induction (Fig. 5A). The induction profile for the four pcp genes in the pcpM mutant was very similar to that in the wild type (Fig. 5B). In contrast to the pcpM mutant, the pcpR mutant failed to express pcpB and pcpE upon PCP induction, and the transcription of pcpA did not increase (Fig. 5C).

FIG. 5.

RT-PCR analysis of transcription of the pcp genes in ATCC 39723 and its mutants. Cells were induced by 75 μM PCP for 1 h at an optical density of about 0.35 and harvested. Uninduced cells were collected at an optical density of about 0.4. Total RNA was extracted, and expression of individual pcp genes was analyzed by RT-PCR. (A) Wild type; (B) pcpM mutant; (C) pcpR mutant. Lanes 1 and 10, 1-kb DNA ladder; lanes 2 to 5, uninduced; lanes 6 to 9, induced; lanes 2 and 6, pcpB (expected size of RT-PCR product, 548 bp); lanes 3 and 7, pcpC (604 bp); lanes 4 and 8, pcpA (518 bp); lanes 5 and 9, pcpE (433 bp).

Identification of inducers.

The pcpB mutant, Fco315, was used to determine whether PCP itself could be an inducer because Fco315 cannot transform PCP (15). Primers BF-1 and BR-1 were located in the N-terminal coding region of pcpB preceding the insertion site of the kanamycin cassette (15) and were used to detect pcpB transcription. Addition of PCP induced the expression of pcpB, pcpA, and pcpE (Fig. 6A). Thus, PCP itself served as an inducer to activate three pcp genes in ATCC 39723. Since PcpB initiates the mineralization of a broad range of polychlorinated phenols, including 2,3,5,6-tetrachlorophenol (TeCP), TriCP, and DiCP (38), these polychlorinated phenols were tested for the ability to be inducers. The pcpB mutant was induced with TeCP, TriCP, or DiCP, and the transcription of pcpB was analyzed. All the chlorinated phenols tested were able to induce transcription of pcpB (Fig. 6B).

FIG. 6.

RT-PCR analysis of induction of the pcp genes by polychlorinated phenols in the pcpB mutant. The mutant cells were induced for 1 h by 75 μM PCP or other polychlorinated phenols and harvested. The total RNA was extracted and used to analyze the induction of the pcp genes. (A) Induction of the pcp genes by PCP. Lanes 1 and 8, 1-kb DNA ladder; lane 2, uninduced pcpB; lanes 3 to 6, PCP-induced pcpB, pcpC, pcpA, and pcpE, respectively; lane 7, total RNA amplified by Taq polymerase with pcpB-specific primers as a negative control. (B) Induction of pcpB by polychlorinated phenols. Lanes 1 and 9, 1-kb DNA ladder; lane 2, total RNA template amplified by Taq polymerase; lane 3, genomic DNA amplified by Taq polymerase as a positive control; lane 4, uninduced; lane 5, PCP induced; lane 6, TeCP induced; lane 7, TriCP induced; lane 8, DiCP induced.

Specific activities of PcpB, PcpC, PcpA, and PcpE in cell extracts.

To validate the RT-PCR results, the specific activities of PcpB, PcpC, PcpA, and PcpE in cell extracts of ATCC 39723 and its mutants were determined (Table 3). The activities were in good agreement with the RT-PCR results. PcpB, PcpA, and PcpE activities were inducible, while PcpC activity was constitutive. PcpB activity was not detectable without induction. PcpA had a basal level of activity, but the activity of this enzyme increased upon PCP induction. Although 2-CMA reductase activity increased significantly upon PCP induction, there were relatively high 2-CMA reductase activities in the uninduced wild-type cells and even in the pcpE mutant cells. We suspect that there is another MA reductase besides PcpE. The induced pcpR mutant had the same levels of PcpB, PcpA, PcpE, and PcpC activities as the uninduced wild type. For the pcpM mutant the overall induction pattern of the four enzymes was similar to that of the wild type, except that the induced PcpA activity was five times higher than that in the wild type.

TABLE 3.

Specific activities of PCP-degrading enzymes in different genetic backgrounds

Straina	Sp act (nmol min−1 mg−1) ofb:
	PcpB	PcpA	PcpE	PcpC
Wild type, PCP induced	0.78 ± 0.04	1.10 ± 0.08	46.28 ± 2.13	7.48 ± 0.33
Wild type, uninduced	0c	0.29 ± 0.04	3.45 ± 0.31	6.41 ± 1.39
pcpR, PCP induced	0	0.24 ± 0.08	4.95 ± 0.45	5.83 ± 0.63
pcpM, PCP induced	0.78 ± 0.07	5.98 ± 0.09	28.47 ± 1.08	6.52 ± 1.55
pcpM, uninduced	0	0.08 ± 0.04	3.20 ± 0.76	6.61 ± 0.39
pcpB, PCP induced	0	2.38 ± 0.20	17.77 ± 0.94	8.47 ± 0.47
pcpA, PCP induced	NTd	0	NT	8.47 ± 0.58
pcpE, PCP induced	NT	NT	3.25 ± 0.20	10.40 ± 0.71
pcpC, PCP induced	NT	NT	NT	0

^aThe pcpR, pcpM, pcpB, pcpA, pcpE, and pcpC strains are mutants in which the pcpR, pcpM, pcpB, pcpA, pcpE, and pcpC genes, respectively, are inactivated.

^bThe values are averages ± standard deviations for at least three measurements.

^cA value of zero indicates that activity was undetectable.

^dNT, not tested.

DISCUSSION

PcpE was shown to be responsible for 2-CMA reduction during PCP degradation in ATCC 39723 (Fig. 1) by two lines of evidence: 2-CMA accumulated only in the pcpE mutant culture (Fig. 4A), and 2-CMA reductase activity increased about 14-fold in the wild type but not in the pcpE mutant after PCP induction (Table 3). Characterization of PcpE as a functional MA reductase was done with recombinant PcpE purified from an E. coli host (3). PcpE reduced 2-CMA to MA and then to 3-oxoadipate by consuming two NADH (3), which is consistent with the reported function of MA reductases in other bacteria (12, 32). Some microorganisms may have two or more isofunctional MA reductases (14, 28). The lack of stoichiometric accumulation of 2-CMA (Fig. 3) and the detectable MA activity (Table 3) in the pcpE mutant suggest that there is another isofunctional MA reductase in ATCC 39723. However, the activity of this isofunctional MA reductase does not increase during PCP degradation, nor is it enough to prevent the accumulation of 2-CMA in the pcpE mutant. Thus, four catabolic gene products, PcpB, PcpC, PcpA, and PcpE, are responsible for converting PCP to 3-oxoadipate (Fig. 1). 3-Oxoadipate, a central metabolic intermediate of many aromatic compounds, is channeled by 3-oxoadipate:succinyl coenzyme A transferase and 3-oxoadipyl coenzyme A thiolase into the tricarboxylic acid cycle for mineralization in bacteria (11).

In this study, the organization of the pcp genes in ATCC 39723 was documented. Five pcp genes involved in PCP degradation have been identified. They are the catabolic genes pcpB, pcpC, pcpA, and pcpE and the regulatory gene pcpR. There are also an uncharacterized pcpD gene (21) and a hypothetical regulatory gene, pcpM. The genes are scattered on two fragments. pcpB, pcpD, and pcpR are organized as a cluster with the same orientation. pcpB and pcpD are possibly cotranscribed. However, pcpR is not likely cotranscribed with pcpBD because a rho-independent terminator sequence is located 38 bp downstream of pcpD. pcpC, pcpA, pcpM, and pcpE are on another fragment, but they are apparently arranged in discrete locations and transcribed from different promoters. Such a discrete organization is uncommon for genes of a single degradation pathway in bacteria. We speculate that the discrete organization may reflect the recent acquisition of the pcp genes in ATCC 39723 for the degradation of PCP, which was first introduced as a wood preservative in 1936 (5).

To examine the possible origins of the pcp genes, the G+C contents and codon usage of the pcp genes and their flanking sequences were examined. The G+C contents are 60.78% for pcpC, 60.96% for pcpA, 62.34% for pcpB, 63.79% for pcpD, 63.96% for pcpR, 64.72% for pcpM, and 66.48% for pcpE. The average G+C content is similar to the local G+C content of the pcpEMAC fragment (63.02%) and the overall G+C content of the ATCC 39723 genome (63.8%) (24). The codon usage of the pcp genes showed the general trend of GC bias at the wobble positions for most amino acids, reflecting the high G+C content. However, TAT (codon usage frequency, 0.67) is favored over TAC (0.33) for tyrosine, and CAT (0.66) is more common than CAC (0.34) for histidine. One of the leucine codons (CTA) was not used in any pcp gene. Comparison of the codon preference between the pcp genes and the adjacent genes did not reveal any apparent differences. Thus, the pcp genes likely evolved from the same organism or related organisms.

The roles of two putative LysR-type regulators in PCP degradation were investigated. Genetic and functional analyses showed that pcpR is required for PCP degradation, while pcpM is not. The complete loss of the ability to degrade PCP (Fig. 4A) and the lack of transcription of pcpB and pcpE in the pcpR mutant (Fig. 5C) suggest that PcpR is an essential activator for these genes. Using the pcpB mutant, we found that pcpB, pcpA, and pcpE are inducible by PCP (Fig. 6A). The induction of pcpB by several other polychlorinated phenols (Fig. 6B) provides evidence that polychlorinated phenols are coinducers of PcpR. In addition, the pcpM mutant degraded PCP almost as well as the wild type, and inactivation of the pcpM gene did not change the expression profile of the pcp genes (Fig. 5B). From these results, we concluded that PcpM is not critical for PCP degradation and that PcpR regulates the expression of pcpB, pcpA, and pcpE.

Conserved binding sites of PcpR in the regulatory regions of pcpB, pcpA, and pcpE were identified. A LysR-type regulator generally binds to a 15-bp region of disrupted dyadic sequence centered near position −65 from the transcriptional start; there is a conserved 5′-T-N₁₁-A-3′ motif termed the LysR motif within the dyadic sequence (10, 26). By using these criteria, a conserved sequence, 5′-ATTC-N₇-GAAT-3′, was identified in the promoter sequences of pcpB, pcpA, and pcpE (Table 4). The predicted binding site is also similar to the ATAC-N₇-GTAT sequence for CatM and BenM binding involved in the regulation of benzoate and catechol degradation in an Acinetobacter sp. (6). The presence of conserved binding sites in the promoter regions of pcpB, pcpA, and pcpE further suggests that these genes are regulated as a regulon with PcpR as the activator and polychlorinated phenols as coinducers. Although the four functional genes responsible for the conversion of PCP to 3-oxoadipate are organized in four discrete locations, three genes are regulated as a regulon with pcpC constitutively expressed.

TABLE 4.

Putative binding sites of LysR-type transcriptional regulators in the promoters of the pcp genes

Regulated gene	Interrupted dyadic sequencea	Location on the promoter
pcpB	5′-ATTCAGGTTCCGAAT-3′	−85 to −71b
pcpA	5′-ATTCATTATTCGAAT-3′	−75 to −57b
pcpE	5′-ATTCGTTCATCGAAT-3′	−173 to −159c
pcpC	5′-ACCGTTCGCGCAGCGAATGGT-3′	−75 to −55b

^aThe interrupted dyadic sequences are indicated by boldface type, and the 5′-T-N₁₁-A-3′ LysR motifs are underlined.

^bLocation relative to the transcriptional start position of pcpB (22), pcpC (20), or pcpA (36).

^cLocation relative to the position of the translational start codon of pcpE.

The deduced amino acid sequences encoded by pcpR and pcpM are most similar to each other; the overall level of identity is 43%, and there is significant conservation throughout the entire polypeptide sequences. The sequence similarity between pcpR and pcpM suggests a common ancestor for these two genes. It is surprising that PcpM is not required for pcpA activation because of the location of pcpM relative to the location of pcpA and the similarity between PcpM and PcpR. However, pcpA expression is quite intriguing. pcpA was expressed at a basal level without PCP induction, and expression of this gene was the highest of the pcp genes upon induction (Fig. 5A). Basal-level expression of pcpA was also observed in the pcpR mutant (Fig. 5C). In addition, the induced PcpA activity was higher in the pcpM mutant than in the wild type (Table 3). Unfortunately, RT-PCR analysis did not show more pcpA expression in the pcpM mutant than in the wild type upon PCP induction (Fig. 5A and B). Since the RT-PCR signal for pcpA is quite strong in both cases, perhaps RT-PCR amplification of pcpA was not in the linear range for induced pcpA. On the basis of the gene expression (Fig. 5) and enzyme activity (Table 3), we speculated that PcpM may compete with PcpR for the pcpA promoter. If PcpM binds to the promoter but does not use PCP as a coinducer, it may reduce the activation by PcpR. This speculation is in agreement with our conclusion that PcpM is not important in PCP degradation. PcpA and PcpM likely evolved from the metabolism of other compounds rather than the metabolism of polychlorinated phenols, and PcpM may respond to other coinducers. These hypotheses are subject to validation by further studies.

Since PcpC is constitutively produced in ATCC 39723, there has been speculation concerning its function in other cellular events (21). We showed that a pcpC mutation did not affect the ability of cells to grow on glutamate in the mineral medium. The pcpC mutant also degraded TriCP and DiCP as well as the wild type did (data not shown). The only effect of pcpC mutation was on the metabolism of TeCH, which transitorily accumulated in the culture and then was further transformed to a product with a 347-nm absorption peak (Fig. 4C). When TeCH was directly added to a pcpC mutant culture or cultures of several unrelated bacteria, it was also transformed into the product with the 347-nm absorption peak (data not shown). This product may be similar to an oxidized derivative of DiCH, which has an absorption peak at 350 nm; the product accumulates in the pcpA mutant after PCP degradation and is likely a polymer of DiCH (4). Therefore, PcpC is the enzyme that metabolizes TeCH during PCP degradation in ATCC 39723. It is unclear why pcpC is constitutively expressed. One possibility is that PcpC has other functions in ATCC 39723 but the functions are not essential. Another possible explanation is that the regulation of pcpC expression has not been evolved in PCP degradation. A potential binding site for a LysR-type regulator was identified in the pcpC promoter sequence (Table 4). Because the sequence is different from the potential PcpR binding sequence, a different LysR-type regulator may have regulated the expression of pcpC. If the regulator cannot respond to PCP or its metabolic intermediates, the next solution would be to make pcpC expression constitutive through spontaneous mutation and selection. Sequence analysis classified PcpC as a zeta-class glutathione S-transferase (20). The zeta-class glutathione S-transferases are known to catalyze the isomerization of maleylacetoacetate to fumarylacetoacetate, which is an essential step in the catabolism of tyrosine and phenylalanine, and the dechlorination of dichloroacetate to glyoxylate (2). Besides dechlorinating TeCH, PcpC also isomerizes maleylacetone (an analog of maleylacetoacetate) and consumes dichloroacetate (1). The biochemical evidence also suggests that pcpC is newly recruited for PCP degradation.

With this study, each gene responsible for the metabolism of PCP to 3-oxoadipate, a common metabolite of many aromatic compounds, has been inactivated, and the roles of the genes in PCP degradation in ATCC 39723 have been demonstrated. Both pcpB and pcpA have been inactivated previously and have been shown to participate in PCP degradation (4, 15). In this study, pcpE was identified and was shown to be responsible for conversion of 2-CMA to 3-oxoadipate during PCP degradation. The role of PcpC in TeCH dechlorination during PCP degradation was confirmed with the pcpC mutant. Thus, the PCP degradation pathway in ATCC 39723 has been characterized both biochemically and genetically (Fig. 1). The unique and discrete organization of each catalytic gene has implied that there may be multiple regulation mechanisms. Instead, a single regulator has been found to control induction of pcpA, pcpB, and pcpE as a regulon. Together with constitutively expressed pcpC, the regulation system is rather straightforward. These findings should facilitate genetic engineering of microorganisms with improved catalytic properties for bioremediation of polychlorinated phenols and their derivatives.